一种基于正负差图像的运动目标检测新方法

New method for moving target detection based on positive and negative difference images

  • 摘要: 运动目标检测领域中现有的差图像法是利用绝对值差图像检测差图像上运动目标区域,用现有方法检测时易受噪声干扰,而且当摄像机有自运动时需要进行背景运动补偿。因此,提出一种新算法,即首先分别计算正差图像与负差图像,然后利用运动目标区域在正差图像与负差图像中的幅值、形状以及运动等信息的对称性对其进行检测,最后给出针对飞机尾焰序列图像进行检测的结果。实验结果表明:该方法可提高运动目标检测的可靠性与效率。

     

    Abstract: Image differencing is a widely used method in moving target detection nowadays. Since most of the work only use the absolute value of difference images and the motion of the background must be compensated, the reliability and practicality of the method are not satisfactory. A new moving target detection method that uses both the positive and negative difference images is proposed in this paper. The new method detects moving target by calculating the positive and negative difference images respectively and making use of the symmetry of amplitude, shape and motion information for a moving target area in positive and negative difference images. The detection results of aircraft tailflame sequential images are given at the end of the paper. The experiment shows that the method can improve the reliability and efficiency for the moving target detection.

     

/

返回文章
返回